1. Technical Field
The present invention is directed to an apparatus and method for error logging on a memory module.
2. Description of Related Art
In computing systems, the ability to detect hardware errors in memory has been available for many years. Recently, a mechanism for collecting memory errors and storing them in the computer system in a format in which the computer system user could identify which Dual In-line Memory Module (DIMM) sockets contain defective DIMMs has been devised.
However, this mechanism has the drawback that there is no connection maintained between the DIMM error data stored on the system and the defective DIMM. That is, the defective DIMM may be removed from the DIMM socket in the motherboard and there will be no way to identify the DIMM as being faulty based on the DIMM error data stored in the system. As a result, the faulty DIMM may be inserted into another DIMM socket on the same or a different computing system and may cause errors.
In order to avoid this drawback, International Business Machines Corporation (IBM) has devised a mechanism for recording the type and quantity of errors caused by a DIMM within a storage device mounted to the DIMM has been devised. This mechanism is described in IBM Research Disclosure 424100, entitled “On Part Memory DIMM Error Logging,” disclosed by IBM in August, 1999. This mechanism, however, does not provide any ability to identify which memory elements on the memory module are faulty. Thus, while the DIMM may be identified as faulty, there is no way to identify which part of the DIMM is faulty without performing a detailed analysis of the DIMM prior to repairing the DIMM.
Furthermore, the information stored on the DIMM in the prior art mechanism is the functional status of the DIMM. This allows for any system to determine the operational condition of the DIMM and whether or not to configure the DIMM for system use. Even if the DIMM were moved to another system, that new system could determine that the DIMM had previously failed and not configure the defective DIMM for system usage. This simple status record does not contain the required additional information required to repair the defective DIMM.
Thus, it would be beneficial to have an apparatus and method for error logging in which the error log is permanently connected with the faulty DIMM and identifies which part of the DIMM is faulty in order to facilitate repair of the DIMM.